Social-LLM: Modeling User Behavior at Scale Using Language Models and Social Network Data
Abstract
1. Introduction
- We propose Social-LLM, a social network representation model designed to be scalable that combines user content cues with social network cues for inductive user detection tasks.
- We conduct a thorough evaluation of Social-LLMs on 7 real-world, large-scale social media datasets of various topics and detection tasks.
- We showcase the utility of using Social-LLM embeddings for visualization.
2. Related Work
3. Social-LLM Framework
3.1. Content Cues
3.2. Network Cues
3.3. User Representation Module
3.4. Unsupervised Training via Siamese Architecture
3.5. Multiple Negatives Ranking Loss
Algorithm 1 Alternating Training Loss for Retweet and Mention Edges |
|
3.6. Downstream Task Application
3.7. Trade-Offs of Social-LLM
4. Experiments
4.1. Data
4.1.1. COVID Politics
4.1.2. Election 2020
4.1.3. COVID Morality
4.1.4. Ukraine–Russia Suspended Accounts
4.1.5. Ukraine–Russia Hate
4.1.6. Immigration Hate
4.2. Evaluation
4.2.1. Content-Based Methods
4.2.2. Network-Based Methods
4.2.3. Hybrid Method
4.2.4. Experimental Setup
4.3. Results
4.3.1. Experiment 1: Choice of LLMs
4.3.2. Experiment 2: Main Experiments
4.3.3. Experiment 3: Edge Type Ablation
4.3.4. Experiment 4: Edge Weights and Directions
4.3.5. Experiment 5: User Tweet Embeddings
4.3.6. Experiment 6: Sensitivity to Dimension Size
4.3.7. Visualization
5. Conclusions
5.1. Limitations
5.2. Ethical Considerations
Author Contributions
Funding
Data Availability Statement
Conflicts of Interest
References
- Freeman, L.C. The Development of Social Network Analysis: A Study in the Sociology of Science; Empirical Press: Vancouver, BC, Canada, 2004. [Google Scholar]
- Lazer, D.; Pentland, A.; Adamic, L.; Aral, S.; Barabási, A.-L.; Brewer, D.; Christakis, N.; Contractor, N.; Fowler, J.; Gutmann, M.; et al. Computational Social Science. Science 2009, 323, 721–723. [Google Scholar] [CrossRef]
- Ma, H.; Rong, Y.; Huang, J. Graph Neural Networks: Scalability. In Graph Neural Networks: Foundations, Frontiers, and Applications; Springer Nature: New York, NY, USA, 2022; pp. 99–119. [Google Scholar]
- Wu, Z.; Pan, S.; Chen, F.; Long, G.; Zhang, C.; Yu, P.S. A Comprehensive Survey on Graph Neural Networks. IEEE Trans. Neural Netw. Learn. Syst. 2021, 32, 4–24. [Google Scholar] [CrossRef] [PubMed]
- McPherson, M.; Smith-Lovin, L.; Cook, J.M. Birds of a Feather: Homophily in Social Networks. Annu. Rev. Sociol. 2001, 27, 415–444. [Google Scholar] [CrossRef]
- Kovacs, B.; Kleinbaum, A.M. Language-Style Similarity and Social Networks. Psychol. Sci. 2020, 31, 202–213. [Google Scholar] [CrossRef] [PubMed]
- Jiang, J.; Ren, X.; Ferrara, E. Retweet-BERT: Political Leaning Detection Using Language Features and Information Diffusion on Social Networks. In Proceedings of the International AAAI Conference on Web and Social Media (ICWSM), Limassol, Cyprus, 5–8 June 2023; Volume 17, pp. 459–469. [Google Scholar]
- Barberá, P.; Jost, J.T.; Nagler, J.; Tucker, J.A.; Bonneau, R. Tweeting From Left to Right: Is Online Political Communication More Than an Echo Chamber? Psychol. Sci. 2015, 26, 1531–1542. [Google Scholar] [CrossRef]
- Jiang, J.; Chen, E.; Yan, S.; Lerman, K.; Ferrara, E. Political Polarization Drives Online Conversations About COVID-19 in the United States. Hum. Behav. Emerg. Technol. 2020, 2, 200–211. [Google Scholar] [CrossRef]
- Davis, C.A.; Varol, O.; Ferrara, E.; Flammini, A.; Menczer, F. Botornot: A System to Evaluate Social Bots. In Proceedings of the 25th International Conference Companion on World Wide Web (WWW Companion), Montreal, QC, Canada, 11–15 May 2016; pp. 273–274, ISBN 9781450341448. [Google Scholar]
- Masood, F.; Almogren, A.; Abbas, A.; Khattak, H.A.; Din, I.U.; Guizani, M.; Zuair, M. Spammer Detection and Fake User Identification on Social Networks. IEEE Access 2019, 7, 68140–68152. [Google Scholar] [CrossRef]
- Jiang, J.; Luceri, L.; Walther, J.B.; Ferrara, E. Social Approval and Network Homophily as Motivators of Online Toxicity. arXiv 2023, arXiv:2310.07779. [Google Scholar] [CrossRef]
- Ribeiro, M.; Calais, P.; Santos, Y.; Almeida, V.; Meira, W., Jr. Characterizing and Detecting Hateful Users on Twitter. In Proceedings of the International AAAI Conference on Web and Social Media (ICWSM), Palo Alto, CA, USA, 25–28 June 2018; Volume 12. [Google Scholar]
- Rios, S.A.; Aguilera, F.; Nuñez-Gonzalez, J.D.; Graña, M. Semantically Enhanced Network Analysis for Influencer Identification in Online Social Networks. Neurocomputing 2019, 326, 71–81. [Google Scholar] [CrossRef]
- Aral, S.; Walker, D. Identifying Influential and Susceptible Members of Social Networks. Science 2012, 337, 337–341. [Google Scholar] [CrossRef]
- Ye, J.; Luceri, L.; Jiang, J.; Ferrara, E. Susceptibility to Unreliable Information Sources: Swift Adoption with Minimal Exposure. In Proceedings of the ACM Web Conference 2024, Singapore, 13–17 May 2024; pp. 4674–4685. [Google Scholar]
- Gopalakrishnan, A.A.; Hossain, J.; Elmas, T.; Sarıyüce, A.E. Density-aware Walks for Coordinated Campaign Detection. arXiv 2025, arXiv:2506.13912. [Google Scholar] [CrossRef]
- Choi, J. User Authentication Using Graph Neural Networks for Adapting to Dynamic and Evolving User Patterns. Electronics 2025, 14, 3570. [Google Scholar] [CrossRef]
- Tzoumanekas, G.; Chatzianastasis, M.; Ilias, L.; Kiokes, G.; Psarras, J.; Askounis, D. A Graph Neural Architecture Search Approach for Identifying Bots in Social Media. Front. Artif. Intell. 2024, 7, 1509179. [Google Scholar] [CrossRef] [PubMed]
- Hamilton, W.L. Graph Representation Learning; Morgan & Claypool Publishers: San Rafael, CA, USA, 2020. [Google Scholar]
- Liu, Z.; Chen, C.; Yang, X.; Zhou, J.; Li, X.; Song, L. Heterogeneous Graph Neural Networks for Malicious Account Detection. In Proceedings of the 27th ACM International Conference on Information and Knowledge Management (CIKM ’18), Turin, Italy, 22–26 October 2018; pp. 2077–2085. [Google Scholar]
- Serafini, M.; Guan, H. Scalable Graph Neural Network Training: The Case for Sampling. ACM SIGOPS Oper. Syst. Rev. 2021, 55, 68–76. [Google Scholar] [CrossRef]
- Xiao, Z.; Song, W.; Xu, H.; Ren, Z.; Sun, Y. TIMME: Twitter Ideology-Detection via Multi-Task Multi-Relational Embedding. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual, 6–10 July 2020; pp. 2258–2268. [Google Scholar]
- Zhang, X.; Liu, Y.; Wang, J.; Hu, Z.; Chen, X.; Yan, R. SAGraph: A Large-Scale Social Graph Dataset with Comprehensive Context for Influencer Selection in Marketing. In Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, Washington, DC, USA, 14–18 July 2024; pp. 3733–3742. [Google Scholar]
- Pasham, S.D. Scalable Graph-Based Algorithms for Real-Time Analysis of Big Data in Social Networks. Metascience 2024, 2, 92–129. [Google Scholar]
- Qiu, Z.; Ma, C.; Wu, J.; Yang, J. Text is All You Need: LLM-enhanced Incremental Social Event Detection. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL), Vienna, Austria, 27 July–1 August 2025; pp. 4666–4680. [Google Scholar]
- Li, L.; Liu, G.; Liu, Y.; Yu, Q.; Luo, C.; Li, N. A Novel Rumor Detection Method Focusing on Social Psychology with Graph Attention Network. Neurocomputing 2025, 626, 129609. [Google Scholar] [CrossRef]
- Thapa, S.; Shiwakoti, S.; Shah, S.B.; Adhikari, S.; Veeramani, H.; Nasim, M.; Naseem, U. Large Language Models (LLM) in Computational Social Science: Prospects, Current State, and Challenges. Soc. Netw. Anal. Min. 2025, 15, 1–30. [Google Scholar] [CrossRef]
- Jiang, J.; Chen, E.; Luceri, L.; Murić, G.; Pierri, F.; Chang, H.-C.H.; Ferrara, E. What are Your Pronouns? Examining Gender Pronoun Usage on Twitter. In Proceedings of the 17th International AAAI Conference on Web and Social Media, Limassol, Cyprus, 5–8 June 2023. [Google Scholar]
- Piao, G.; Breslin, J.G. Inferring User Interests for Passive Users on Twitter by Leveraging Followee Biographies. In Proceedings of the 39th European Conference on IR Research, ECIR 2017, Aberdeen, UK, 8–13 April 2017; pp. 122–133. [Google Scholar]
- Rogers, N.; Jones, J.J. Using Twitter Bios to Measure Changes in Self-Identity: Are Americans Defining Themselves More Politically Over Time? J. Soc. Comput. 2021, 2, 1–13. [Google Scholar] [CrossRef]
- Thelwall, M.; Thelwall, S.; Fairclough, R. Male, Female, and Nonbinary Differences in UK Twitter Self-Descriptions: A Fine-Grained Systematic Exploration. J. Data Inf. Sci. 2021, 6, 1–27. [Google Scholar] [CrossRef]
- Martha, V.; Zhao, W.; Xu, X. A Study on Twitter User-Follower Network: A Network Based Analysis. In Proceedings of the 2013 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining, Niagara Falls, ON, Canada, 25–28 August 2013; pp. 1405–1409. [Google Scholar]
- Conover, M.; Ratkiewicz, J.; Francisco, M.; Gonçalves, B.; Menczer, F.; Flammini, A. Political Polarization on Twitter. In Proceedings of the International AAAI Conference on Web and Social Media (ICWSM), Barcelona, Spain, 17–21 July 2011; Volume 5, pp. 89–96. [Google Scholar]
- Ferrara, E.; Varol, O.; Davis, C.; Menczer, F.; Flammini, A. The Rise of Social Bots. Commun. ACM 2016, 59, 96–104. [Google Scholar] [CrossRef]
- Boyd, D.; Golder, S.; Lotan, G. Tweet, Tweet, Retweet: Conversational Aspects of Retweeting on Twitter. In HICSS 2010; IEEE: Piscataway, NJ, USA, 2010; pp. 1–10. [Google Scholar]
- Metaxas, P.; Mustafaraj, E.; Wong, K.; Zeng, L.; O’Keefe, M.; Finn, S. What do Retweets Indicate? Results From User Survey and Meta-Review of Research. In Proceedings of the International AAAI Conference on Web and Social Media (ICWSM), Oxford, UK, 26–29 May 2015; Volume 9, pp. 658–661. [Google Scholar]
- Hemsley, J.; Stromer-Galley, J.; Semaan, B.; Tanupabrungsun, S. Tweeting to the Target: Candidates’ Use of Strategic Messages and @Mentions on Twitter. J. Inf. Technol. Politics 2018, 15, 3–18. [Google Scholar] [CrossRef]
- Reimers, N.; Gurevych, I. Sentence-BERT: Sentence Embeddings Using Siamese BERT-Networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), Hong Kong, China, 3–7 November 2019. [Google Scholar]
- Henderson, M.; Al-Rfou, R.; Strope, B.; Sung, Y.-H.; Lukács, L.; Guo, R.; Kumar, S.; Miklos, B.; Kurzweil, R. Efficient Natural Language Response Suggestion for Smart Reply. arXiv 2017, arXiv:1705.00652. [Google Scholar] [CrossRef]
- Calvillo, D.P.; Ross, B.J.; Garcia, R.J.B.; Smelter, T.J.; Rutchick, A.M. Political Ideology Predicts Perceptions of the Threat of COVID-19 (and Susceptibility to Fake News About It). Soc. Psychol. Personal. Sci. 2020, 11, 1119–1128. [Google Scholar] [CrossRef]
- Chen, E.; Lerman, K.; Ferrara, E. Tracking Social Media Discourse About the COVID-19 Pandemic: Development of a Public Coronavirus Twitter Data Set. JMIR Public Health Surveill. 2020, 6, e19273. [Google Scholar] [CrossRef] [PubMed]
- Yang, K.-C.; Ferrara, E.; Menczer, F. Botometer 101: Social Bot Practicum for Computational Social Scientists. J. Comput. Soc. Sci. 2022, 5, 1511–1528. [Google Scholar] [CrossRef] [PubMed]
- Chen, E.; Deb, A.; Ferrara, E. #Election2020: The First Public Twitter Dataset on the 2020 US Presidential Election. J. Comput. Soc. Sci. 2022, 5, 1–18. [Google Scholar]
- Haidt, J.; Joseph, C. Intuitive Ethics: How Innately Prepared Intuitions Generate Culturally Variable Virtues. Daedalus 2004, 133, 55–66. [Google Scholar] [CrossRef]
- Haidt, J. The Righteous Mind: Why Good People Are Divided by Politics and Religion; Vintage: New York, NY, USA, 2012. [Google Scholar]
- Rao, A.; Guo, S.; Wang, S.-Y.N.; Morstatter, F.; Lerman, K. Pandemic Culture Wars: Partisan Asymmetries in the Moral Language of COVID-19 Discussions. In Proceedings of the 2023 IEEE International Conference on Big Data (BigData), Sorrento, Italy, 15–18 December 2023; pp. 413–422. [Google Scholar]
- Dehghani, M.; Johnson, K.; Hoover, J.; Sagi, E.; Garten, J.; Parmar, N.J.; Vaisey, S.; Iliev, R.; Graham, J. Purity Homophily in Social Networks. J. Exp. Psychol. Gen. 2016, 145, 366. [Google Scholar] [CrossRef]
- Chan, E.Y. Moral Foundations Underlying Behavioral Compliance During the COVID-19 Pandemic. Personal. Individ. Differ. 2021, 171, 110463. [Google Scholar] [CrossRef]
- Díaz, R.; Cova, F. Reactance, Morality, and Disgust: The Relationship Between Affective Dispositions and Compliance With Official Health Recommendations During the COVID-19 Pandemic. Cogn. Emot. 2022, 36, 120–136. [Google Scholar] [CrossRef]
- Francis, K.B.; McNabb, C.B. Moral Decision-Making During COVID-19: Moral Judgements, Moralisation, and Everyday Behaviour. Front. Psychol. 2022, 12, 6484. [Google Scholar] [CrossRef]
- Hoover, J.; Portillo-Wightman, G.; Yeh, L.; Havaldar, S.; Davani, A.M.; Lin, Y.; Kennedy, B.; Atari, M.; Kamel, Z.; Mendlen, M.; et al. Moral Foundations Twitter Corpus: A Collection of 35k Tweets Annotated for Moral Sentiment. Soc. Psychol. Personal. Sci. 2020, 11, 1057–1071. [Google Scholar] [CrossRef]
- Indiana University’s Observatory on Social Media. Suspicious Twitter Activity around the Russian Invasion of Ukraine. 2022. Available online: https://osome.iu.edu/research/white-papers/Ukraine_OSoMe_White_Paper_March_2022.pdf (accessed on 29 September 2025).
- Pierri, F.; Luceri, L.; Jindal, N.; Ferrara, E. Propaganda and Misinformation on Facebook and Twitter during the Russian Invasion of Ukraine. In Proceedings of the 15th ACM Web Science Conference 2023 (WebSci), Austin, TX, USA, 30 April–1 May 2023; pp. 65–74, ISBN 9798400700897. [Google Scholar]
- Chen, E.; Ferrara, E. Tweets in Time of Conflict: A Public Dataset Tracking the Twitter Discourse on the War Between Ukraine and Russia. In Proceedings of the International AAAI Conference on Web and Social Media (ICWSM), Limassol, Cyprus, 5–8 June 2023; Volume 17, pp. 1006–1013. [Google Scholar]
- Pierri, F.; Luceri, L.; Chen, E.; Ferrara, E. How Does Twitter Account Moderation Work? Dynamics of Account Creation and Suspension on Twitter During Major Geopolitical Events. EPJ Data Sci. 2023, 12, 43. [Google Scholar] [CrossRef] [PubMed]
- Thapa, S.; Shah, A.; Jafri, F.A.; Naseem, U.; Razzak, I. A Multi-Modal Dataset for Hate Speech Detection on Social Media: Case-Study of Russia-Ukraine Conflict. In Proceedings of the 5th Workshop on Challenges and Applications of Automated Extraction of Socio-political Events from Text (CASE), Abu Dhabi, United Arab Emirates, 7–8 December 2022. [Google Scholar]
- Frimer, J.A.; Aujla, H.; Feinberg, M.; Skitka, L.J.; Aquino, K.; Eichstaedt, J.C.; Willer, R. Incivility is Rising Among American Politicians on Twitter. Soc. Psychol. Personal. Sci. 2023, 14, 259–269. [Google Scholar] [CrossRef]
- Kim, J.W.; Guess, A.; Nyhan, B.; Reifler, J. The Distorting Prism of Social Media: How Self-Selection and Exposure to Incivility Fuel Online Comment Toxicity. J. Commun. 2021, 71, 922–946. [Google Scholar] [CrossRef]
- Bianchi, F.; HIlls, S.; Rossini, P.; Hovy, D.; Tromble, R.; Tintarev, N. “It’s Not Just Hate”: A Multi-Dimensional Perspective on Detecting Harmful Speech Online. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing (EMNLP), Abu Dhabi, United Arab Emirates, 7–11 December 2022; pp. 8093–8099. [Google Scholar]
- Liu, Y.; Ott, M.; Goyal, N.; Du, J.; Joshi, M.; Chen, D.; Levy, O.; Lewis, M.; Zettlemoyer, L.; Stoyanov, V. RoBERTa: A Robustly Optimized BERT pretraining approach. arXiv 2019, arXiv:1907.11692. [Google Scholar]
- Nguyen, D.Q.; Vu, T.; Nguyen, A.T. BERTweet: A Pre-Trained Language Model for English Tweets. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP): System Demonstrations, Online, 16–20 November 2020; pp. 9–14. [Google Scholar]
- Song, K.; Tan, X.; Qin, T.; Lu, J.; Liu, T.-Y. MPNet: Masked and Permuted Pre-Training for Language Understanding. Adv. Neural Inf. Process. Syst. 2020, 33, 16857–16867. [Google Scholar]
- Grover, A.; Leskovec, J. Node2vec: Scalable Feature Learning for Networks. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), San Francisco, CA, USA, 13—17 August 2016; pp. 855–864. [Google Scholar]
- Zhang, J.; Dong, Y.; Wang, Y.; Tang, J.; Ding, M. Prone: Fast and Scalable Network Representation Learning. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence (IJCAI-19), Macao, China, 10–16 August 2019; Volume 19, pp. 4278–4284. [Google Scholar]
- Hamilton, W.; Ying, Z.; Leskovec, J. Inductive Representation Learning on Large Graphs. Adv. Neural Inf. Process. Syst. 2017, 1025–1035. [Google Scholar]
- Bosley, M.; Jacobs-Harukawa, M.; Licht, H.; Hoyle, A. Do We Still Need BERT in the Age of GPT? Comparing the Benefits of Domain-Adaptation and In-Context-Learning Approaches to Using LLMs for Political Science Research. In Proceedings of the 2023 Annual Meeting of the Midwest Political Science Association (MPSA), Chicago, IL, USA, 13–16 April 2023. [Google Scholar]
- Van der Maaten, L.; Hinton, G. Visualizing Data Using t-SNE. J. Mach. Learn. Res. 2008, 9, 2579–2605. [Google Scholar]
- Ferrara, E. Fairness and bias in artificial intelligence: A brief survey of sources, impacts, and mitigation strategies. Sci 2024, 6, 3. [Google Scholar] [CrossRef]
# RT | # MN | Profile | Metadata | Tweet | Time | Pred. | |||
---|---|---|---|---|---|---|---|---|---|
Dataset | # Users | Edges | Edges | Desc. | Features | Texts | Span | Label(s) | Type |
1. Covid-Political | 78,672 | 180,928 | - | ✓ | ✓ | ✗ | 6 Mo | Partisanship (1) | Cls. |
2. Election2020 | 78,932 | 2,818,603 | - | ✓ | ✗ | ✗ | 3 Mo | Partisanship (1) | Cls. |
3. COVID-Morality | 119,770 | 609,845 | 639,994 | ✓ | ✓ | ✗ | 2 Yr | Morality (5) | Reg. |
4. Ukr-Rus-Suspended | 56,440 | 135,053 | 255,476 | ✓ | ✓ | ✓ | 1 Mo | Suspension (1) | Cls. |
5. Ukr-Rus-Hate | 82,041 | 166,741 | 414,258 | ✓ | ✓ | ✗ | 1 Mo | Toxicity (6) | Reg. |
6. Immigration-Hate-08 | 5759 | 63,097 | 83,870 | ✓ | ✓ | ✗ | All | Toxicity (5) | Reg. |
7. Immigration-Hate-05 | 2188 | 4827 | 7993 | ✓ | ✓ | ✗ | All | Toxicity (5) | Reg. |
Election | Covid- | Ukr-Rus- | Covid- | Ukr-Rus- | Immigration-Hate- | ||||
---|---|---|---|---|---|---|---|---|---|
2020 | Political | Suspended | Morality | Hate | 05 | 08 | |||
C | N | Cls. (Metric: Macro-F1) | Reg. (Metric: Pearson) | ||||||
Experiment 1: LLMs | |||||||||
RoBERTa | ✓ | ✗ | 80.11 | 78.41 | 56.21 | 32.84 | 36.54 | 12.06 | 9.30 |
BERTweet | ✓ | ✗ | 79.31 | 78.42 | 55.69 | 30.72 | 40.38 | 14.33 | 12.03 |
SBERT-MPNet | ✓ | ✗ | 86.47 | 82.99 | 56.79 | 36.77 | 43.35 | 17.16 | 16.76 |
Experiment 2 (Main): Baselines vs. Social-LLM | |||||||||
(a) Profile LLM | ✓ | ✗ | 86.47 | 82.99 | 56.79 | 36.77 | 43.35 | 17.16 | 16.76 |
(a) + (b) Metadata | ✓ | ✗ | - | 83.26 | 70.75 | 40.43 | 45.38 | 17.72 | 17.32 |
(a) + (b) + (c) Tweet LLMs | ✓ | ✗ | - | - | 81.74 | - | - | - | - |
(d) node2vec | ✗ | ✓ | - | 88.65 | 72.33 | 50.53 | 39.97 | 10.70 | 12.18 |
(e) ProNE | ✗ | ✓ | 76.28 | 64.04 | 77.95 | 51.13 | 45.38 | 5.47 | 14.30 |
(f) TIMME | ✓ | ✓ | 84.81 | 81.85 | 72.91 | 30.47 | 43.46 | 20.98 | 18.67 |
Social-LLM | ✓ | ✓ | 97.87 | 90.82 | 82.71 | 50.15 | 57.27 | 21.17 | 20.11 |
%↑ | 13% | 2% | 1% | −2% | 26% | 1% | 7% | ||
Experiment 3: Ablation on edge types in Social-LLM models | |||||||||
RT | ✓ | ✓ | - | - | 70.71 | 46.57 | 48.18 | 18.85 | 18.18 |
MN | ✓ | ✓ | - | - | 71.32 | 45.33 | 49.55 | 18.73 | 17.92 |
RT & MN (distinct) | ✓ | ✓ | - | - | 71.99 | 20.40 | 51.20 | 14.75 | 18.89 |
RT + MN (indistinct) | ✓ | ✓ | - | - | 72.10 | 47.51 | 50.73 | 19.05 | 18.53 |
Experiment 4: Ablation on edge directions and weights in Social-LLM models | |||||||||
(best edge combo model) | ✓ | ✓ | 97.78 | 90.68 | 72.10 | 47.51 | 50.73 | 18.53 | 19.05 |
+ w | ✓ | ✓ | 97.78 | 90.55 | 71.85 | 46.98 | 51.46 | 18.70 | 18.81 |
+ d | ✓ | ✓ | 97.85 | 90.82 | 71.77 | 50.15 | 57.19 | 18.95 | 18.77 |
+ d + w | ✓ | ✓ | 97.82 | 90.42 | 72.17 | 46.89 | 49.35 | 17.67 | 19.21 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Jiang, J.; Ferrara, E. Social-LLM: Modeling User Behavior at Scale Using Language Models and Social Network Data. Sci 2025, 7, 138. https://doi.org/10.3390/sci7040138
Jiang J, Ferrara E. Social-LLM: Modeling User Behavior at Scale Using Language Models and Social Network Data. Sci. 2025; 7(4):138. https://doi.org/10.3390/sci7040138
Chicago/Turabian StyleJiang, Julie, and Emilio Ferrara. 2025. "Social-LLM: Modeling User Behavior at Scale Using Language Models and Social Network Data" Sci 7, no. 4: 138. https://doi.org/10.3390/sci7040138
APA StyleJiang, J., & Ferrara, E. (2025). Social-LLM: Modeling User Behavior at Scale Using Language Models and Social Network Data. Sci, 7(4), 138. https://doi.org/10.3390/sci7040138